AI CV & Job Search System
Project Lead / AI Workflow Lead · Working prototype · Ongoing
A working prototype that uses AI agents to review and improve CVs, discover jobs from multiple sources and generate tailored cover letters.
AI CV & Job Search System is a working prototype that helps job seekers improve their CV, find more relevant roles across multiple sources and generate tailored cover letters.
The opportunity
Most candidates apply to many roles with one CV. ATS systems filter on keywords, while hiring managers look for relevance. A structured, multi-agent workflow can bridge that gap without removing human judgement.
My role
I designed the workflow, validation criteria and user journey, then mapped how each agent contributes — CV improvement, job discovery, cover-letter generation, diagnostics and quality checks — before building the orchestration logic that moves a user from CV upload to application-ready output.
Key work completed
- CV upload and parsing (PDF or TXT, up to 10 MB).
- CV quality report and improved CV preview with structured output.
- Profile confirmation with target market, preferred locations, job titles and keywords.
- Generic CV vs per-job tailored CV mode.
- Multi-source job discovery with relevance scoring.
- Source tags for LinkedIn, Greenhouse, Lever, myPOS, Flow, Nuvei and others.
- Select all / select top controls (10, 25, 50, 100).
- Per-job cover-letter generation with approve and download-all options.
Results & current status
The prototype is functional and being refined. No public demo URL or broader user numbers have been confirmed.
Tools & methods
- Multi-agent AI workflow design
- Prompt engineering
- ATS and recruitment domain knowledge
- Process mapping
- Quality gates
- API orchestration
- Python
- Windsurf
Lessons learned
Output quality depends on input quality and review steps. A well-structured prompt chain beats a single heroic prompt, and a human check at the end prevents embarrassing mistakes.